1: % Prepared by Karl Ecklund September 2002

2: % Questions, Improvements, Comments to kme and CLEOAC

3: %

4: % LATEX 2e Template for CLEO Papers

5: % YOU MUST USE REVTEX4 and latex 2e

6: %

7: % Checklist for Paper Drafts:

8: % ---------------------------

9: % 0) Use appropriate \documetclass line as indicated below

10: % 1) Draft Number: Use latest CBX number, append A for the first vote

11: %                                (B,C,... for subsequent if any votes)

12: % 2) Don't use CLNS or CLEO numbers - this happens after your vote.

13: % 3) Title; use \\ to break title over several lines.

14: % 4) Abstract

15: % 5) For the Author list use CLEO Collaboration only

16: % 6) Body

17: % 

18: % Checklist for Journal Submissions:

19: % ----------------------------------

20: % 0) Use appropriate \documetclass line as indicated below

21: %    - For CLNS, hep-ex, preprints, conf. paper use CLNS version

22: %    - For PRL submission use PRL version

23: %    - For PRD submission use PRD version 

24: %      AND use \author{(CLEO Collaboration)} instead of \collaboration{CLEO}

25: %      SEE PRD_SPECIAL_CHANGEME in author list during step 5 below

26: % 1) CLEO paper number (from CLEOAC)

27: % 2) CLNS preprint number (from CLEOAC)

28: % 3) Title; use \\ to break title over several lines.

29: % 4) Abstract

30: % 5) Author list (from CLEOAC)

31: % 6) PACS codes

32: % 7) Body

33: % 8) Add acknowledgments

34: % 9) Hardcode the \date when ready to submit to journal and hep-ex.

35: %

36: % Checklist for Conference Papers:

37: % --------------------------------

38: % 0) Use appropriate \documetclass line as indicated below

39: %    - For conf. paper use CLNS version

40: % 1) CLEO conference paper number (from CLEOAC) (don't use CLNS)

41: % 2) Title; use \\ to break title over several lines.

42: % 3) To \thanks after title, add appropriate conference info.

43: % 4) Abstract

44: % 5) Author list (from CLEOAC)

45: % 6) Body

46: % 7) Add acknowledgments

47: % 8) Hardcode the \date when ready to submit to conference and hep-ex.

48: % 

49: % KNOWN PROBLEMS with template and REVTEX4

50: % - can't have \collaboration in PRD style grouped author list

51: %   Using \author{(CLEO Collaboration)} instead

52: % - can't make abstract appear before full author list ala CLNS notes

53: %   Abandoning this as the format for CLNS.

54: 

55: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

56: % Select one of the \documentclass lines below for your paper

57: %%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%%

58: %%%%%%%%%%%%%% Use for CLNS preprint (hep-ex) and Paper Drafts

59: \documentclass[aps,prd,preprint,superscriptaddress,tightenlines,nofootinbib,floatfix]{revtex4}

60: 

61: %%%%%%%%%%%%%% Use for PRL

62: %\documentclass[aps,prl,twocolumn,superscriptaddress,showpacs]{revtex4}

63: 

64: %%%%%%%%%%%%%% Use for PRD submission

65: %\documentclass[aps,prd,preprint,nopreprintnumbers,nofootinbib,showpacs]{revtex4}

66: %%%%%%%%%%%%%% Use for PRD formatting tables and figures in 2 column

67: %\documentclass[aps,prd,twocolumn,nofootinbib,showpacs]{revtex4}

68: 

69: \usepackage{graphicx}% Include figure files

70: \usepackage{dcolumn}% Align table columns on decimal point

71: \usepackage{bm}% bold math

72: \usepackage{multirow}% multirow table entries

73: 

74: \begin{document}

75: 

76: %-Definitions-----------------------------------------------------------

77: \newcommand{\barb}{\bar{B}}

78: %\newcommand{\bbbar}{B\bar{B}}

79: %\newcommand{\bm}{B^-}

80: \newcommand{\bob}{\bar{B}^0}

81: \newcommand{\bo}{B^0}

82: \newcommand{\bp}{B^+}

83: \newcommand{\dkpipiz}{D^0 \to K^-\pi^+\pi^0}

84: \newcommand{\dkpi}{D^0 \to K^-\pi^+}

85: \newcommand{\dkrho}{D^0 \to K^- \rho^+}

86: \newcommand{\dobar}{\overline{D}^0}

87: \newcommand{\etal}{{\it et al.}}

88: \newcommand{\F}{$\cal F$}

89: %\newcommand{\gev}{\ \rm GeV}

90: \newcommand{\kpipiz}{K^-\pi^+\pi^0}

91: %\newcommand{\kpi}{K^-\pi^+}

92: \newcommand{\krho}{K^- \rho^+}

93: %\newcommand{\mev}{\ \rm MeV}

94: \newcommand{\piz}{\pi^0}

95: \newcommand{\ra}{{\rightarrow}}

96: \newcommand{\Ufs}{\Upsilon(4S)}

97: 

98: \newcommand{\about}		{\mbox{$\sim$}}

99: \newcommand{\aerr}[3]   {\mbox{${{#1}^{+ #2}_{- #3}}$}}

100: \newcommand{\amp}    	{\mbox{${\cal A}$}}

101: \newcommand{\avcb}		{\mbox{$|V_{cb}|$}}

102: \newcommand{\avtb}		{\mbox{$|V_{tb}|$}}

103: \newcommand{\avtd}		{\mbox{$|V_{td}|$}}

104: \newcommand{\avts}		{\mbox{$|V_{ts}|$}}

105: \newcommand{\avub}		{\mbox{$|V_{ub}|$}}

106: \newcommand{\avud}		{\mbox{$|V_{ud}|$}}

107: \newcommand{\avus}		{\mbox{$|V_{us}|$}}

108: \newcommand{\BBbar}     {\mbox{$B\bar B$}}

109: \newcommand{\bbbar}     {\mbox{$B\bar B$}}

110: \newcommand{\bc}		{\mbox{$b\to c$}}

111: \newcommand{\bdb}		{\mbox{$\bar B^0_d $}}

112: \newcommand{\bd}		{\mbox{$B^0_d $}}

113: \newcommand{\berr}[2]   {\mbox{${{}^{+ #1}_{- #2}}$}}

114: \newcommand{\bmeson}	{\mbox{$B$}}

115: \newcommand{\bqb}		{\mbox{$\bar B^0_q $}}

116: \newcommand{\bq}		{\mbox{$B^0_q $}}

117: \newcommand{\branch}    {\mbox{${\cal B}$}}

118: \newcommand{\bsb}		{\mbox{$\bar B^0_s$}}

119: \newcommand{\bs}		{\mbox{$B^0_s$}}

120: \newcommand{\bu}		{\mbox{$b\to u$}}

121: \newcommand{\cerr}[4]   {\mbox{${{}^{+ #1}_{- #2}{}^{+ #3}_{- #4}}$}}

122: \newcommand{\chisq}		{\mbox{$\chi^2$}}

123: \newcommand{\cosB}		{\mbox{${\cos\theta_{\rm B}}$}}

124: \newcommand{\cossph}	{\mbox{$\cos \theta_{\rm sph}$}}

125: \newcommand{\costhr}	{\mbox{$\cos \theta_{\rm thr}$}}

126: \newcommand{\dedx}		{\mbox{$dE/dx$}}

127: \newcommand{\derr}[5]   {\mbox{${{#1}^{+ #2}_{- #3}{}^{+ #4}_{- #5}}$}}

128: \newcommand{\de}		{\mbox{$\Delta E$}}

129: \newcommand{\dzerobar}	{\mbox{$\overline {D^0}$}}

130: \newcommand{\dzero}		{\mbox{${D^0}$}}

131: \newcommand{\ebeam}		{\mbox{$E_{\rm beam}$}}

132: \newcommand{\eb}		{\mbox{$E_b$}}

133: \newcommand{\eeqq}		{\mbox{$e^+e^-\to\qqb$}}

134: \newcommand{\ee}		{\mbox{$e^+e^-$}}

135: %\newcommand{\etal}		{\mbox{${\it et al}$}}

136: \newcommand{\expt}		{\mbox{$_{\rm expt}$}}

137: \newcommand{\fbinv}		{\mbox{${\rm fb}^{-1}$}}

138: %\newcommand{\fisher}    {\mbox{${\cal F}$}}

139: \newcommand{\fisher}    {\mbox{$x_{\cal F}$}}

140: \newcommand{\gev}		{\mbox{${\rm ~GeV}$}}

141: \newcommand{\hh}		{\mbox{$h^+h^-$}}

142: \newcommand{\hpm}		{\mbox{$h^\pm$}}

143: \newcommand{\implies}	{\mbox{${\Longrightarrow}$}}

144: \newcommand{\jimexp}[1]	{\mbox{${\rm e}^{{#1}}$}}

145: \newcommand{\kk}		{\mbox{KK}}

146: \newcommand{\Kpi}		{\mbox{$K\pi$}}

147: \newcommand{\kpi}		{\mbox{$\Kpi$}}

148: \newcommand{\kpz}		{\mbox{$K^+\pi^0$}}

149: \newcommand{\KP}		{\mbox{$K\pi$}}

150: \newcommand{\ksp}		{\mbox{$K^0_S\pi^+$}}

151: \newcommand{\ks}		{\mbox{$K^0_S$}}

152: \newcommand{\kz}		{\mbox{$K^0$}}

153: \newcommand{\kzb}		{\mbox{$\overline{K^0}$}}

154: \newcommand{\like}    	{\mbox{${\cal L}$}}

155: \newcommand{\Lp}		{\mbox{$\Lambda \bar p$}}

156: \newcommand{\lum}    	{\mbox{${\cal L}$}}

157: \newcommand{\mb}		{\mbox{$M_B$}}

158: \newcommand{\mev}		{\mbox{${\rm MeV}$}}

159: \newcommand{\micron}	{\mbox{$~\mu{\rm m}$}}

160: \newcommand{\mm}		{\mbox{$\mu^+\mu^-$}}

161: \newcommand{\model}		{\mbox{$_{\rm mod}$}}

162: \newcommand{\nbb}		{\mbox{$N_{B\bar B}$}}

163: \newcommand{\nbinv}		{\mbox{${\rm nb}^{-1}$}}

164: \newcommand{\pb}        {\mbox{$p_B$}}

165: \newcommand{\pbinv}		{\mbox{${\rm pb}^{-1}$}}

166: \newcommand{\pdf}		{\mbox{${PDF}$}}

167: \newcommand{\pdfs}		{\mbox{${PDF}{\rm s}$}}

168: \newcommand{\pidkpi}	{\mbox{$\Delta_{K\pi}$}}

169: \newcommand{\pidkp}		{\mbox{$\Delta_{Kp}$}}

170: \newcommand{\pid}		{\mbox{$\Delta_{PID}$}}

171: \newcommand{\pipi}		{\mbox{$\Pipi$}}

172: \newcommand{\Pipi}		{\mbox{$\pi\pi$}}

173: \newcommand{\power}[1]  {\mbox{${\times 10^{#1}}$}}

174: \newcommand{\ppz}		{\mbox{$\pi^+\pi^0$}}

175: \newcommand{\pvec}		{\mbox{$\vec{p}$}}

176: \newcommand{\pz}		{\mbox{$\pi^0$}}

177: \newcommand{\qqb}		{\mbox{$q\bar q$}}

178: \newcommand{\qq}		{\mbox{${q\bar q}$}}

179: \newcommand{\stat}		{\mbox{$_{\rm stat}$}}

180: \newcommand{\syst}		{\mbox{$_{\rm syst}$}}

181: \newcommand{\theo}		{\mbox{$_{\rm theo}$}}

182: \newcommand{\upsi}		{\mbox{$\Upsilon$({\rm 4S})}}

183: \newcommand{\vcb}		{\mbox{$V_{cb}$}}

184: \newcommand{\vcd}		{\mbox{$V_{cd}$}}

185: \newcommand{\vcs}		{\mbox{$V_{cs}$}}

186: \newcommand{\vtb}		{\mbox{$V_{tb}$}}

187: \newcommand{\vtd}		{\mbox{$V_{td}$}}

188: \newcommand{\vts}		{\mbox{$V_{ts}$}}

189: \newcommand{\vub}		{\mbox{$V_{ub}$}}

190: \newcommand{\vud}		{\mbox{$V_{ud}$}}

191: \newcommand{\vus}		{\mbox{$V_{us}$}}

192: \newcommand{\zhat}		{\mbox{$\hat{\bf z}$}}

193: 

194: \newcommand{\ppbar}		{\mbox{${p\bar p}$}}

195: \newcommand{\pL}		{\mbox{${p\bar\Lambda}$}}

196: \newcommand{\LL}		{\mbox{${\Lambda\bar\Lambda}$}}

197: 

198: \newcommand{\mpdf}    	{\mbox{${\cal M}$}}

199: \newcommand{\epdf}    	{\mbox{${\cal E}$}}

200: \newcommand{\fpdf}    	{\mbox{${\cal F}$}}

201: \newcommand{\cpdf}    	{\mbox{${\cal C}$}}

202: \newcommand{\dkmode}	{\mbox{${\mu}$}}

203: \newcommand{\contrib}	{\mbox{${\kappa}$}}

204: \newcommand{\mc}    	{\mbox{${}_{\dkmode\contrib}$}}

205: \newcommand{\nk}    	{\mbox{${n}_{\contrib}$}}

206: 

207: \newcommand{\G}		{\mbox{${G}$}}

208: \newcommand{\GG}	{\mbox{${\cal G}$}}

209: \newcommand{\ARG}	{\mbox{${A}$}}

210: \newcommand{\LIN}	{\mbox{${L}$}}

211: \newcommand{\BW}	{\mbox{${\cal R}$}}

212: \newcommand{\FI}	{\mbox{${F_0}$}}

213: \newcommand{\DG}	{\mbox{${a\G_1+b\G_2}$}}

214: \newcommand{\GGG}	{\mbox{${a\G+b\GG}$}}

215: \newcommand{\DGG}	{\mbox{${a\GG_1+b\GG_2}$}}

216: 

217: \newcommand{\sba} 	{\mbox{${S/B}$}}

218: \newcommand{\rfw}   {\mbox{$R_2$}}

219: \newcommand{\effmc}     {\mbox{${\epsilon_{\rm MC}}$}}

220: \newcommand{\effdata}     {\mbox{${\epsilon_{\rm DATA}}$}}

221: 

222: \newcommand{\RR}	{\mbox{${\cal R}$}}

223: 

224: 

225: 

226: 

227: 

228: 

229: 

230: 

231: 

232: 

233: 

234: 

235: %-----------------------------------------------------------------------

236: %-----------------------------------------------------------------------

237: %-----------------------------------------------------------------------

238: 

239: %\preprint line(s) will be ignored for PRL/PRD

240: %\preprint{CLEO Draft YY-NNA} % For paper draft CBX YY-NN -> Draft YY-NNA

241: %\preprint{CLEO CONF YY-NN}   % For conference papers

242: %\preprint{ICHEP ABSnnn}      % For conference papers

243: \preprint{CLNS 03/1816}       % for CLNS notes

244: \preprint{CLEO 03-03}         % for CLNS notes

245: 

246: \title{Measurements of Charmless Hadronic Two-Body B Meson Decays 

247: and the Ratio $\branch(B\to DK)/\branch(B\to D\pi)$}

248: 

249: %-------- INSERT HERE ------------

250: % Your author list goes here  REMOVE EVERYTHING to END INSERT and

251: % replace with your authorlist (ask cleoac).

252: 

253: 

254: \author{A.~Bornheim}

255: \author{E.~Lipeles}

256: \author{S.~P.~Pappas}

257: \author{A.~Shapiro}

258: \author{W.~M.~Sun}

259: \author{A.~J.~Weinstein}

260: \affiliation{California Institute of Technology, Pasadena, California 91125}

261: \author{R.~A.~Briere}

262: \author{G.~P.~Chen}

263: \author{T.~Ferguson}

264: \author{G.~Tatishvili}

265: \author{H.~Vogel}

266: \affiliation{Carnegie Mellon University, Pittsburgh, Pennsylvania 15213}

267: \author{N.~E.~Adam}

268: \author{J.~P.~Alexander}

269: \author{K.~Berkelman}

270: \author{F.~Blanc}

271: \author{V.~Boisvert}

272: \author{D.~G.~Cassel}

273: \author{P.~S.~Drell}

274: \author{J.~E.~Duboscq}

275: \author{K.~M.~Ecklund}

276: \author{R.~Ehrlich}

277: \author{R.~S.~Galik}

278: \author{L.~Gibbons}

279: \author{B.~Gittelman}

280: \author{S.~W.~Gray}

281: \author{D.~L.~Hartill}

282: \author{B.~K.~Heltsley}

283: \author{L.~Hsu}

284: \author{C.~D.~Jones}

285: \author{J.~Kandaswamy}

286: \author{D.~L.~Kreinick}

287: \author{A.~Magerkurth}

288: \author{H.~Mahlke-Kr\"uger}

289: \author{T.~O.~Meyer}

290: \author{N.~B.~Mistry}

291: \author{J.~R.~Patterson}

292: \author{D.~Peterson}

293: \author{J.~Pivarski}

294: \author{S.~J.~Richichi}

295: \author{D.~Riley}

296: \author{A.~J.~Sadoff}

297: \author{H.~Schwarthoff}

298: \author{M.~R.~Shepherd}

299: \author{J.~G.~Thayer}

300: \author{D.~Urner}

301: \author{T.~Wilksen}

302: \author{A.~Warburton}

303: \author{M.~Weinberger}

304: \affiliation{Cornell University, Ithaca, New York 14853}

305: \author{S.~B.~Athar}

306: \author{P.~Avery}

307: \author{L.~Breva-Newell}

308: \author{V.~Potlia}

309: \author{H.~Stoeck}

310: \author{J.~Yelton}

311: \affiliation{University of Florida, Gainesville, Florida 32611}

312: \author{K.~Benslama}

313: \author{B.~I.~Eisenstein}

314: \author{G.~D.~Gollin}

315: \author{I.~Karliner}

316: \author{N.~Lowrey}

317: \author{C.~Plager}

318: \author{C.~Sedlack}

319: \author{M.~Selen}

320: \author{J.~J.~Thaler}

321: \author{J.~Williams}

322: \affiliation{University of Illinois, Urbana-Champaign, Illinois 61801}

323: \author{K.~W.~Edwards}

324: \affiliation{Carleton University, Ottawa, Ontario, Canada K1S 5B6 \\

325: and the Institute of Particle Physics, Canada M5S 1A7}

326: \author{D.~Besson}

327: \author{X.~Zhao}

328: \affiliation{University of Kansas, Lawrence, Kansas 66045}

329: \author{S.~Anderson}

330: \author{V.~V.~Frolov}

331: \author{D.~T.~Gong}

332: \author{Y.~Kubota}

333: \author{S.~Z.~Li}

334: \author{R.~Poling}

335: \author{A.~Smith}

336: \author{C.~J.~Stepaniak}

337: \author{J.~Urheim}

338: \affiliation{University of Minnesota, Minneapolis, Minnesota 55455}

339: \author{Z.~Metreveli}

340: \author{K.K.~Seth}

341: \author{A.~Tomaradze}

342: \author{P.~Zweber}

343: \affiliation{Northwestern University, Evanston, Illinois 60208}

344: \author{S.~Ahmed}

345: \author{M.~S.~Alam}

346: \author{J.~Ernst}

347: \author{L.~Jian}

348: \author{M.~Saleem}

349: \author{F.~Wappler}

350: \affiliation{State University of New York at Albany, Albany, New York 12222}

351: \author{K.~Arms}

352: \author{E.~Eckhart}

353: \author{K.~K.~Gan}

354: \author{C.~Gwon}

355: \author{K.~Honscheid}

356: \author{D.~Hufnagel}

357: \author{H.~Kagan}

358: \author{R.~Kass}

359: \author{T.~K.~Pedlar}

360: \author{E.~von~Toerne}

361: \author{M.~M.~Zoeller}

362: \affiliation{Ohio State University, Columbus, Ohio 43210}

363: \author{H.~Severini}

364: \author{P.~Skubic}

365: \affiliation{University of Oklahoma, Norman, Oklahoma 73019}

366: \author{S.A.~Dytman}

367: \author{J.A.~Mueller}

368: \author{S.~Nam}

369: \author{V.~Savinov}

370: \affiliation{University of Pittsburgh, Pittsburgh, Pennsylvania 15260}

371: \author{J.~W.~Hinson}

372: \author{J.~Lee}

373: \author{D.~H.~Miller}

374: \author{V.~Pavlunin}

375: \author{B.~Sanghi}

376: \author{E.~I.~Shibata}

377: \author{I.~P.~J.~Shipsey}

378: \affiliation{Purdue University, West Lafayette, Indiana 47907}

379: \author{D.~Cronin-Hennessy}

380: \author{A.L.~Lyon}

381: \author{C.~S.~Park}

382: \author{W.~Park}

383: \author{J.~B.~Thayer}

384: \author{E.~H.~Thorndike}

385: \affiliation{University of Rochester, Rochester, New York 14627}

386: \author{T.~E.~Coan}

387: \author{Y.~S.~Gao}

388: \author{F.~Liu}

389: \author{Y.~Maravin}

390: \author{R.~Stroynowski}

391: \affiliation{Southern Methodist University, Dallas, Texas 75275}

392: \author{M.~Artuso}

393: \author{C.~Boulahouache}

394: \author{S.~Blusk}

395: \author{K.~Bukin}

396: \author{E.~Dambasuren}

397: \author{R.~Mountain}

398: \author{H.~Muramatsu}

399: \author{R.~Nandakumar}

400: \author{T.~Skwarnicki}

401: \author{S.~Stone}

402: \author{J.C.~Wang}

403: \affiliation{Syracuse University, Syracuse, New York 13244}

404: \author{A.~H.~Mahmood}

405: \affiliation{University of Texas - Pan American, Edinburg, Texas 78539}

406: \author{S.~E.~Csorna}

407: \author{I.~Danko}

408: \affiliation{Vanderbilt University, Nashville, Tennessee 37235}

409: \author{G.~Bonvicini}

410: \author{D.~Cinabro}

411: \author{M.~Dubrovin}

412: \author{S.~McGee}

413: \affiliation{Wayne State University, Detroit, Michigan 48202}

414: %\author{(CLEO Collaboration)} %FOR PRD_SPECIAL_CHANGEME

415: \collaboration{CLEO Collaboration} %FOR PRL,CLNS

416: \noaffiliation

417: 

418: %\author{John Doe}

419: %\affiliation{Physics Department, Cornell University

420: %Ithaca, New York 14853}

421: %\author{(CLEO Collaboration)}     %FOR PRD_SPECIAL_CHANGEME

422: %\collaboration{CLEO Collaboration} %FOR PRL and CLNS (superscriptaddress)

423: %\noaffiliation

424: 

425: %-------- END INSERT ------------

426: 

427: %please hard code the date when you have a final draft and submit to CLEOAC

428: \date{\today}

429: 

430: 

431: 

432: %---------------------------------------------------------------------

433: %

434: %\abstract

435: %

436: %---------------------------------------------------------------------

437: 

438: \begin{abstract} 

439: We present final measurements of thirteen charmless hadronic \bmeson\ decay modes from the

440: CLEO experiment. The decay modes include the ten \pipi, \Kpi, and \kk\

441: final states and new limits on dibaryonic final states, \ppbar, \pL, and

442: \LL, as well as a new determination of the ratio $\branch(B\to

443: DK)/\branch(B\to D\pi)$. The results are based on the full CLEO II and

444: CLEO III data samples totalling 15.3 \fbinv\ at the $\Upsilon(4S)$, and

445: supercede previously published results.

446: %comment: cleoiii = 6.18 ON, 2.24 OFF, cleoii = 9.13 ON, 4.35 OFF

447: 

448: \end{abstract}

449: \pacs{13.20.He}

450: \maketitle

451: 

452: %---------------------------------------------------------------------

453: %

454: \section{Introduction}

455: %

456: %---------------------------------------------------------------------

457: 

458: Charmless decays of \bmeson\ mesons may proceed by $b\to u$, $b\to s$,

459: or $b\to d$ transitions.  The latter two mechanisms require flavor

460: changing neutral currents which are not present at tree level in the

461: Standard Model, and therefore must occur through higher order processes

462: such as the penguin mechanism.  Such processes involve loops, which can

463: open the window for particles and physics outside the Standard Model.

464: Even in the absence of such new physics, interference among competing

465: amplitudes for a given decay mode can be exploited to measure CKM

466: phases. There is a significant body of literature\cite{fleischeretal} on

467: the use of the branching ratio $\branch(B\to K^-\pi^+)$ and other

468: charmless modes to determine or constrain the CKM angle $\gamma$, the

469: phase of \vub\ in conventional representations of the CKM matrix.

470: Compared to the methods of extracting $\gamma$ that are based the $B\to DK$

471: decay modes\cite{DK}, these approaches based on charmless decay modes are less

472: clean theoretically, but more promising experimentally because the

473: event yields are significantly higher. 

474: 

475: Recent work on two-body charmless decay modes suggests that the

476: unitarity triangle may be constructable entirely from charmless modes,

477: without recourse to the traditional constraints involving \bmeson\

478: mixing measurements, CP asymmetry in $B\to J/\psi \ks$, or CP violation

479: in kaon decays. The charmless modes therefore offer an independent

480: approach to probe CP violating effects in heavy quark decay. Significant

481: disagreement between these two approaches, if found in experimental

482: results, would directly challenge the Standard Model and its fundamental

483: statement that all CP violating phenomena stem from a single phase in

484: the CKM matrix. Early results based on current data are already

485: available\cite{neubertnew}, and do indicate a degree of inconsistency. In

486: this paper we present new experimental data on charmless modes and note

487: that these data enhance rather than ameliorate the discrepancy.

488: 

489: CLEO has previously published several papers\cite{previous} reporting

490: measurements of charmless hadronic \bmeson\ meson decay modes, including

491: searches for charmless baryonic final states, with the data of the CLEO

492: II experiment.  Here we report corresponding measurements in the new

493: CLEO III data with results for three $\pi\pi$ modes, $B\to \pi^+\pi^-,~

494: \pi^+\pi^0,~\pz\pz$, four $K\pi$ modes, $B\to K^+\pi^-,~K^+\pi^0,~

495: \kz\pi^+,~\kz\pi^0$, three $K\overline{K}$ modes, $B\to K^+K^-,

496: ~\kz K^-,~\kz\kzb$, and three dibaryonic modes, $B\to

497: \ppbar,~\pL,~\LL$. We also merge CLEO II and CLEO III results to

498: determine a final measurement for each mode based on the full CLEO data

499: set, which hereby supercedes our previous publications.  Recent

500: measurements from BABAR and Belle are in excellent agreement with

501: ours\cite{babarandbelle}. We also report a new measurement of the ratio

502: of branching ratios, $\branch(B\to D^0 K^-)/\branch(B\to D^0\pi^-)$.  

503: 

504: Here and throughout this paper charge conjugate modes are implied. We

505: also make use of the notation $h^\pm$ to represent a charged hadron that may be

506: either a kaon or pion.

507: 

508: 

509: %---------------------------------------------------------------------

510: %

511: \section{The CLEO Detector and Datasets}

512: %

513: %---------------------------------------------------------------------

514: CLEO is a general purpose solenoidal magnet detector operating at the

515: Cornell Electron Storage Ring (CESR). The latter is a symmetric-energy

516: storage ring tuned for the data sets discussed here to provide center of

517: mass energies near the \upsi. At $\sqrt{s}=M_{\upsi}$ the hadronic cross

518: section is approximately 4 nb, with 1 nb of $\ee\to\upsi\to\bbbar$ and 3

519: nb of four-flavor continuum \eeqq.  In the CLEO III running period, July

520: 2000 through June 2001, we obtained an integrated luminosity of 6.18

521: \fbinv\ at the \upsi\ and 2.24 \fbinv\ off-resonance, {\it i.e.,} just

522: below the \bbbar\ threshold.  The off-resonance data are used for

523: background determinations.  The on-resonance data corresponds to $\nbb =

524: (5.73\pm 0.47)\power{6}$ \upsi\ decays. The corresponding numbers for

525: the CLEO II running period (1990-1999) are 9.13 \fbinv\ (($9.66\pm

526: 0.19)\power{6} $\upsi\ decays) and 4.35 \fbinv.  Differences in the \nbb\

527: yield per unit integrated luminosity reflect differences in run

528: conditions.

529: 

530: The CLEO III detector\cite{cleo3det} differs from the CLEO II

531: detector\cite{cleo2det} most notably in the inclusion of a ring-imaging

532: Cherenkov device (RICH)\cite{rich} which provides particle

533: identification at all momenta above the Cherenkov threshold. Even at the

534: highest momenta relevant for \bmeson\ physics, about 2.8 \gev, the RICH

535: separates kaons and pions by 2.3 standard deviations. Measurements of

536: specific ionization (\dedx) in the drift chamber provide an additional

537: 2.0 standard deviation separation at the highest momenta. Charged

538: particle tracking is done by the 47-layer drift chamber and a four-layer

539: silicon tracker which reside in a 1.5T solenoidal magnetic field and

540: provide momentum resolution described by $(\sigma_p/p)^2 = A^2 + B^2

541: p^2$ with $A \approx 0.005$ and $B\approx 0.001~\gev^{-1}$. The

542: absolute momentum calibration is confirmed by comparing the invariant

543: mass of standard decays $J/\psi \to \mu^+\mu^-$, $D^0 \to K^-\pi^+$ with

544: PDG values\cite{pdg}. Photons are detected using a 7800-crystal CsI(Tl)

545: electromagnetic

546: calorimeter%\cite{emcal} 

547: which is unchanged between CLEO II and CLEO III.

548: 

549: %---------------------------------------------------------------------

550: %

551: \section{Elements of the analysis}

552: %

553: %---------------------------------------------------------------------

554: The \upsi\ is produced at rest in the lab frame and decays with low $Q$

555: value to a pair of \bmeson\ mesons that travel non-relativistically,

556: with $\pb \sim 300~\mev$.  In this analysis we assume equal rates of

557: $B^+B^-$ and $B^0 \bar{B^0}$ production\cite{fplusminuscleo}. All decay

558: modes studied in this paper are two-body or quasi-two-body modes. Apart

559: from the modest $\pm 150~\mev$ doppler shifts resulting from the motion

560: of the \bmeson\ mesons in the lab frame, the daughter particles are

561: nearly monochromatic, and, up to resolution smearing, jointly carry the

562: full beam energy \eb\ and have invariant mass equal to the \bmeson\ mass

563: \mb. The approximate monochromaticity of the daughters simplifies

564: particle identification and energy resolution, and helps keep the

565: associated systematic errors low. The other \bmeson\ in the event decays

566: into, on average, five charged and five neutral particles, distributed

567: uniformly in the detector acceptance. The principal background to the

568: analysis comes from the non-$b$ hadronic data, \eeqq, with $\qq = u\bar

569: u, ~d\bar d, ~s\bar s, ~{\rm and}~ c\bar c.$ High momentum, back-to-back

570: particles are typical in such events, and some have invariant masses and

571: total energies close to or in the signal region of the \bmeson\ events. 

572: Fortunately, distinctive event topologies separate most of these

573: background events from the signal.  

574: 

575: This analysis has two principal parts: (a) the application of hard 

576: selection criteria to obtain

577: signal-like events, based on kinematics, event topology, and

578: particle identification; (b) the application of an unbinned extended

579: maximum likelihood fit to the surviving event ensembles to extract the

580: yields of signal and background(s) for each mode. The likelihood fit

581: allows us to make maximum use of available information, while avoiding

582: efficiency losses that further selection criteria would entail.

583: 

584: For the purposes of reconstruction, the CLEO III dataset reported on

585: here is divided into two subsets of roughly equal integrated luminosity,

586: which we will call Set A and Set B. The distinction has ultimately no

587: significant effect on the results, but because of changes in event

588: reconstruction algorithms between the two sets, there are slight

589: differences in resolutions and efficiencies -- mostly affecting modes

590: with charged particles -- that we treat separately until the final CLEO

591: III results are reassembled at the end. We provide in Table

592: \ref{tab:oldnew} some informative comparisons between Set A and Set B.

593: 

594: \begin{table}

595: \begin{center}

596: \caption{Features of Set A and Set B.}

597: \smallskip

598: \begin{tabular}{lcc}

599: \hline

600: Quantity & ~~~~Set A & ~~~~Set B\cr 

601: \hline\hline

602: Fraction of total \nbb\     & 55\%    & 45\%  \cr

603: Track Resolution            &         &       \cr

604: \hfil{`A' Coefficient}      & 0.0055  & 0.0044\cr

605: \hfil{`B' Coefficient ($\gev^{-1}$)}      & 0.0011  & 0.0010\cr

606: $K^+\pi^-$ Mode             &         &       \cr

607: \hfil{$\sigma_{\mb}$ (\mev)}  & 2.7     & 2.7   \cr

608: \hfil{$\sigma_{\de}$(\mev)}   & 22      & 19    \cr

609: \hfil{Efficiency}           & 38\%    & 45\%  \cr

610: $K^+\pi^0$ Mode             &         &       \cr

611: \hfil{$\sigma_{\mb}$ (\mev)}  & 3.1     & 3.1   \cr

612: \hfil{$\sigma_{\de}$(\mev)}   & 31      & 31    \cr

613: \hfil{Efficiency}           & 33\%    & 35\%  \cr

614: $\pi^0\pi^0$ Mode\footnote{Resolutions are given as average of low-side and high-side half-resolutions.}           &         &       \cr

615: \hfil{$\sigma_{\mb}$ (\mev)}  & 3.6     & 3.6  \cr

616: \hfil{$\sigma_{\de}$(\mev)}   & 43      & 43    \cr

617: \hfil{Efficiency}           & 22\%    & 22\%  \cr

618: \hline

619: \end{tabular}

620: \label{tab:oldnew}

621: \end{center}

622: \end{table}

623: 

624: 

625: 

626: Charged track and photon candidates are required to satisfy loose

627: quality requirements which reject poorly determined candidates while

628: retaining high efficiency for real tracks and showers. \ks\ candidates

629: are selected from pairs of charged tracks forming well-measured

630: displaced vertices with a $\pi^+\pi^-$ invariant mass within three

631: standard deviations of the nominal \ks\ mass. In addition the vertex

632: must satisfy $|r_{VTX}| > 5$mm in the transverse plane, and

633: ${\vec{p}_{\ks}}\cdot {\vec{r}_{VTX}} > 0 $. The $\ks\to\pz\pz$ mode is

634: not used.  $\Lambda$ candidates consist of $p\pi$ pairs with invariant

635: mass within three standard deviations  of the nominal $\Lambda$ mass.

636: Pairs of photons with an invariant mass within 2.5 standard deviations

637: of the nominal $\pi^0$ mass are kinematically fit with the mass

638: constrained to the nominal $\pi^0$ mass.

639: 

640: 

641: 

642: %---------------------------------------------------------------------

643: %

644: \subsection{General Event Selection}

645: %

646: %---------------------------------------------------------------------

647: Candidates for rare $B$ decay events are selected for further analysis

648: on the basis of two kinematic variables and one event-shape variable.

649: For each candidate, we construct the beam constrained $B$ candidate mass $M_B =

650: \sqrt{(E_b^2-{\vec{p}}^{~2})}$ where $E_b$ is the beam energy, and

651: $\vec{p}$ is the momentum of the candidate computed from the vector sum

652: of the daughter momenta. For real \bmeson\ mesons $|\vec{p}| \ll E_b$

653: and the width of this distribution is dominated by the $\sim 2.5~\mev$

654: intrinsic beam energy spread. The beam energy is determined run by run

655: from CESR lattice information, and slight corrections are applied

656: afterward to ensure that the observed $B^-$ mass in $B^-\to D^0\pi^-$

657: events matches the accepted value\cite{pdg}. In addition we compute the

658: energy balance variable $\Delta E$ = $E - E_b$ where $E$ is the sum of

659: the daughter energies.  The width of this distribution is about 20 \mev\

660: in all charged modes, as determined by the momentum resolution of the

661: tracking systems, and is about 40 \mev\ in modes involving neutral pions.

662: 

663: 

664: Any candidate with $|\Delta E|<400$ MeV and $M_B > 5.2$ GeV is

665: kept. An additional requirement on \cossph, the cosine of the angle

666: between the sphericity axis of the candidate and the sphericity axis of

667: the rest of the event\cite{previous}, is used to reject the dominant

668: \eeqq\ background. All candidates must satisfy the requirement

669: $|\cossph|<0.8$, which rejects approximately 80\% of the background

670: while retaining nearly 80\% of the signal.

671: 

672: 

673: 

674: 

675: 

676: %---------------------------------------------------------------------

677: %

678: \subsection{Particle Identification Requirements (CLEO III)}

679: %

680: %---------------------------------------------------------------------

681: In the case of a candidate mode involving one or more charged pions or

682: kaons, such as $B\to \Kpi$ or $B\to \pi \pz$, each charged track must be

683: positively identified as $K$ or $\pi$.   The pattern of Cherenkov photon

684: hits in the RICH detector is fit to both a kaon and pion hypothesis,

685: each with its own likelihood $\like_K$ and $\like_\pi$.  The mean number

686: of photon hits entering the fit is twelve, and we require a minimum of

687: four.  Calibrated \dedx\ information from the drift chamber is used to

688: compute a \chisq\ for kaon and pion hypotheses.  The RICH and \dedx\

689: results are combined to form an effective \chisq\ difference, 

690: \begin{equation} 

691: \pidkpi = -2\ln\like_K + 2\ln\like_\pi + \chisq_K - \chisq_\pi. 

692: \label{eq:pidkpi-def}

693: \end{equation} 

694: Kaons are identified by $\pidkpi < \delta_K$ and pions by $-\pidkpi <

695: \delta_\pi$, with values of $\delta_K$ and $\delta_\pi$ chosen to yield

696: $(90\pm 3)$\% efficiency as determined in an independent study of tagged

697: kaons and pions obtained from the decay $D^{*+}\to\pi^+D^0~(D^0\to

698: K^-\pi^+)$.  With this choice of $\delta_K$ and $\delta_\pi$, the

699: misidentification rate for kaons faking pions (pions faking kaons) is

700: 11\%(8\%) at momenta around 2.6 \gev. 

701: 

702: For candidate modes involving protons, positive proton identification is

703: required.  \dedx\ does not distinguish well between protons and kaons at

704: the $\sim 2.5~\gev$ momenta of interest, however, so the proton-kaon

705: separation is achieved with a discriminant based only on RICH

706: information: $\pidkp = -2\ln\like_ p+ 2\ln\like_K < \delta_p$. In this case

707: $\delta_p$ is chosen to yield proton (antiproton) identification

708: efficiency of $76\pm 1 (72\pm 1)$\% with a kaon fake rate of 1\%, as

709: determined in an independent study using tagged kaons from the $D^{*+}$

710: sample as above, and protons from $\Lambda \to p\pi$ decays.

711: 

712: 

713: %---------------------------------------------------------------------

714: %

715: \subsection{Event Selection for CLEO II modes}

716: %

717: %---------------------------------------------------------------------

718: We present three results for which we also analyzed the full CLEO II

719: data set, namely $B\rightarrow \kz\kzb$, $B\rightarrow

720: \Lambda\overline{\Lambda}$, and $B\rightarrow p \overline{\Lambda}$. The

721: $\ks$ selection required that the $\ks$ vertex is separated from the

722: beam spot by more that 3 sigma (5.5 sigma for CLEO II.V for which the

723: innermost drift chamber was replaced with a 3 layer silicon vertex

724: detector). The candidate mass must lie within 10 MeV of the nominal

725: $\ks$ mass. We require that the $\ks$ flight direction points to

726: within 3 sigma of the beamspot .

727: 

728: The protons in the $\pL$ final state must be compatible

729: within 3 sigma with a proton \dedx\ hypothesis and incompatible 

730: with both the electron (calculated from calorimeter information) and muon

731: (calculated from muon chamber information) hypotheses. We require that 

732: the $\Lambda$ candidate mass lie within 10 MeV of the nominal $\Lambda$

733: mass, the vertex be at least 5 mm removed radially from the beam spot,

734: and the $\chi^2$ of the vertex fit be less than nine. 

735: There is no particle identification applied to daughter particles of the

736: $\Lambda$ decays.

737: 

738: 

739: 

740: %---------------------------------------------------------------------

741: %

742: \section{Analysis Variables}

743: %

744: %---------------------------------------------------------------------

745: Events which meet all the requirements described in the preceding

746: paragraphs are now used in a likelihood fit to extract signal yield.

747: We characterize each candidate event by four variables: the mass and

748: energy variables introduced above, \mb\ and \de, the flight direction

749: of the candidate \bmeson, and a Fisher discriminant\cite{Fisher}. 

750: 

751: The flight direction is given by $\cosB = \hat{\bf p}\cdot \zhat$

752: where ${\bf p}$ is the vector sum of the daughter momenta and \zhat\ is

753: the beam axis. Since the vector \upsi\ is produced in \ee\ annihilation

754: it has a polarization $J_z=\pm 1$, and the subsequent

755: flight direction of the pseudoscalar \bmeson\ mesons 

756: is distributed as $|Y^{\pm 1}_1(\theta,\phi)|^2\sim\sin^2\theta = 1-\cos^2\theta$. 

757: Background events are flat in this variable.

758: 

759: The Fisher discriminant is used to refine the separation of signal and

760: \eeqq\ background that is initially addressed by the hard cut on

761: \cossph\ in the general event selection. The Fisher discriminant,

762: \fisher, is a linear combination of fourteen variables, with

763: coefficients chosen to maximize the separation of signal and background

764: events.  The optimization procedure uses Monte Carlo events for the

765: signal and off-resonance and (\mb,\de) sideband data events for the

766: background. As in previous CLEO publications\cite{previous} the

767: component variables include the direction of the thrust

768: axis of the candidate with respect to the beam axis,

769: \costhr, and the nine conical bins of a ``Virtual Calorimeter" whose axis

770: is aligned with the candidate thrust axis. A fuller description of the

771: Virtual Calorimeter is available in a previous 

772: publication\cite{bigrareb}. Note that \costhr\ and \cosB\ are quite

773: different quantities.  For two body decay $B\to XY$, one has simple

774: closed form expressions:  $\cosB = \hat{\bf p}\cdot \zhat$

775: with ${\bf p}\equiv{\bf p}_X+{\bf p}_Y$, whereas 

776: $\costhr  = \hat{\bf q}\cdot \zhat$ with ${\bf q}\equiv{\bf p}_X-{\bf p}_Y$.

777: 

778: In addition, we take advantage of the high quality particle

779: identification in CLEO III to augment the Fisher discriminant with

780: information on the presence of electrons, muons, protons, and kaons in

781: the event. The momentum of the highest momentum electron, muon, kaon,

782: and proton are used as inputs to the Fisher discriminant.  For these

783: purposes we need only rudimentary particle identification criteria. If

784: any of the possible particle type 

785: hypotheses has no corresponding track identified (which is very often the

786: case), a value zero is used as the input to the Fisher discriminant. 

787: %Any

788: %track that passes the simple electron identification criteria is called

789: %an electron;  any remaining track that passes the simple muon

790: %identification criteria is called a muon; any remaining track that is

791: %less than 1 \gev\ and is within three sigma of the proton \dedx\

792: %hypothesis is called a proton; any remaining track that is less than 750

793: %\mev\ and is within three sigma of the kaon or pion \dedx\ hypothesis is

794: %called a kaon or pion, respectively; any remaining track that is

795: %identified by the RICH detector as being more likely to be a kaon than a

796: %pion is called a kaon. And finally all remaining tracks are called pions. 

797: 

798: The Fisher variable thus defined provides discrimination between 

799: charmless \bmeson\ decay signal modes and \qq\ background at a level

800: equivalent to two gaussian distributions separated by 1.4$\sigma$, 

801: and is independent of the details of the signal mode for all the modes

802: studied here.  

803: %In Fig. \ref{fig:fish} we show \qq\ background data and 

804: %Monte Carlo events together with $B\to D^0\pi$ data and Monte Carlo to

805: %illustrate both the separation of signal and background, and the quality

806: %of agreement between data and Monte Carlo. 

807: 

808: 

809: %---------------------------------------------------------------------

810: %

811: \section{Likelihood Fit}

812: %

813: %---------------------------------------------------------------------

814: %---------------------------------------------------------------------

815: %

816: \subsection{Fit Components}

817: %

818: %---------------------------------------------------------------------

819: With the four analysis variables \mb, \de, \fisher, and \cosB, we

820: characterize each event in terms of normalized probability distribution

821: functions (\pdfs): $\mpdf\mc(\mb)$, $\epdf\mc(\de)$,

822: $\fpdf\mc(\fisher)$, and $\cpdf\mc(\cosB)$. The thirteen different

823: charmless decay modes to be fit

824: will in general have contributions from (a) signal, (b) \qq\ background,

825: and (c) cross-feed from other \bmeson\ modes. Subscripts \dkmode\ and

826: \contrib\ identify the particular decay mode (\dkmode) and the type of

827: contribution (\contrib). The probability that a given event characterized

828: by (\mb, \de, \fisher, \cosB) is an event of component type \contrib\ of

829: decay mode \dkmode\ is then given by the product of \pdfs 

830: 

831: \begin{equation}

832: %P\mc(\mb, \de, \fisher, \cosB)=

833: P\mc=

834: \mpdf\mc(\mb)\epdf\mc(\de)\fpdf\mc(\fisher)\cpdf\mc(\cosB).

835: \label{eq:evtprob}

836: \end{equation}

837: 

838: We determine the yields \nk\ of signal, \qq\ background, 

839: and cross-feed background in decay mode \dkmode\ by maximizing

840: the extended likelihood function with respect to the yields \nk:

841: 

842: \begin{equation}

843: \like_{\dkmode} (n_{sig}, n_{\qq}, n_{xfeed})= {\exp(-{ \sum_{\contrib} {\nk}})

844: \prod_{\rm events}{\left( \sum_{\contrib} {\nk P\mc}\right)}.}

845: \label{eq:likelihood}

846: \end{equation}

847: 

848: The \qq\ background is the dominant background source in all cases, and

849: in only five of the fifteen modes do we need to include any cross-feed

850: backgrounds.  Four of these are due to the $\sim 10\%$ $K/\pi$

851: misidentification probability.  In fitting $B\to\pi^+\pi^-$ and

852: $B\to\pi^+\pi^0$ we include components for $B\to K^+\pi^-$ and $B\to

853: K^+\pi^0$, respectively;  in $B\to K^+ K^-$ we include a component for

854: $B\to K^+\pi^-$; and in $B\to D^0 K^-$ we include a component for $B\to

855: D^0\pi^-$. Although the cross-feed backgrounds arise from mistaken

856: particle identification, they are still distinguishable from the signal

857: through $\epdf(\de)$, which is shifted by about 50 \mev\ relative to the

858: signal \pdf. The cross-feed fits are only for background removal and the

859: yields are not used in any other signal determination.

860: 

861: 

862: The fifth mode requiring a cross-feed component is $B^0\to\pz\pz$. In

863: this case a small contribution from $B^+\to\rho^+\pi^0$ arises when the

864: charged pion has very little momentum in the lab frame. Although the

865: missing particle also shifts \de\ by at least one pion mass, resolution

866: smearing leaves a small tail in the signal region. The treatment here is

867: the same as in our previous publication on

868: $B^0\to\pz\pz$\cite{previous}. We note also that potential feedthrough

869: of $B^+\to\rho^+\pi^0$ into $B^+\to \pi^+\pz$ is smaller than in the

870: \pz\pz\ case because the low-side resolution smearing is less for the

871: $\pi^+\pz$ mode, and because the ratio $\branch(B^+\to

872: \pi^+\pz)/\branch(B^+\to\rho^+\pi^0)$ is larger than $\branch(B^+\to

873: \pz\pz)/\branch(B^+\to\rho^+\pi^0)$. Monte Carlo studies confirm these

874: observations and we therefore do not include this term in the $\pi^+\pz$

875: fit.

876: 

877: 

878: 

879: %---------------------------------------------------------------------

880: %

881: \subsection{\pdfs}\label{sec:pdfs}

882: %

883: %---------------------------------------------------------------------

884: We parametrize the \pdfs\ with various functions and combinations of

885: functions which are listed below. In each case the parameters of the

886: function are determined from a fit to signal Monte Carlo event samples

887: for the signal component and cross-feed component (if there is one), and

888: from a fit to off-resonance data for the \qq\ background component. 

889: These parameters are then fixed for all subsequent fitting procedures so

890: the only free variables in the likelihood function Eq.

891: \ref{eq:likelihood} are the signal and background yields.  There is of

892: course underlying uncertainty in the parameter values which fix the

893: \pdf\ shapes, but this uncertainty is systematic in nature and will be

894: discussed later in section \ref{sec:syst}. All functions are normalized

895: to unit area over the accepted range of the free variable.

896: 

897: \begin{itemize}

898: 

899: \item Gaussian (\G): 

900: used for \mb\ and \de\ signal component \pdfs\ that do not involve

901: neutral pions.

902: The parameters are the mean and width.   

903: 

904: \item Asymmetric Gaussian (\GG): 

905: used for \mb\ and \de\ in modes where neutral pions appear. Fluctuations in the

906: measured \pz\ energy are intrinsically asymmetric -- with a longer tail

907: on the low energy side -- because of energy leakage out the back of the

908: CsI crytals in the electromagnetic calorimeter. The parameters are the

909: mean and separate left and right widths.

910: 

911: \item Linear (\LIN): 

912: used for \qq\ backgrounds in \de.  The free parameter is the slope.

913: 

914: \item ARGUS (\ARG): 

915: used to characterize the \mb\ shape of \qq\

916: backgrounds\cite{argus}. $\ARG(\mb) =

917: \sqrt{1-\mb^2/E_b^2}\exp(-\lambda(1-\mb^2/E_b^2))$. The parameter

918: $\lambda$ governs the turn-over of the shape and the slope at low values

919: of \mb. The beam energy $E_b$ determines the endpoint of the \qq\ \mb\

920: spectrum.  Over the course of CLEO III data taking this end point

921: clusters around several close but not identical values.  In practice we

922: form a sum of three ARGUS functions with different endpoint values,

923: weighted by the corresponding integrated luminosities. In addition, to

924: account for run-to-run beam energy variation, we

925: convolve each ARGUS function with a Gaussian of width $\sigma \sim

926: 0.7~\mev$ in $E_b$.

927: 

928: \item Breit-Wigner (\BW): 

929: used in Fisher parametrizations to describe

930: non-Gaussian tails. Parameters are mean and width.

931: 

932: \item Fisher (\FI): 

933: a linear combination of functions used to

934: characterize the \qq\ background Fisher shape.  

935: It is primarily  an asymmetric Gaussian (87\% of the area), but includes an

936: additional Breit-Wigner (9\%) with the same mean, and a small symmetric Gaussian (4\%). 

937: \end{itemize}

938: 

939: Table \ref{tab:allfits} lists the \pdfs\ used for each fit component of

940: each mode.  The fourth fit variable, \cosB\ is in all cases taken to have

941: the functional form $1-\cos^2\theta_{\rm B}$ for signal 

942: and cross-feed components, and flat for \qq\ background.

943: 

944: \begin{table}

945: \begin{center}

946: \caption{Functional forms used in likelihood fits. See text for

947: discussion of terms. Linear combinations are indicated by

948: coefficients $a$ and $b$.}

949: \smallskip

950: \begin{tabular}{llccc}

951: \hline

952: Mode & Fit Component & \mpdf(\mb) & \epdf(\de) & \fpdf(\fisher)  \cr 

953: \hline\hline 

954: $\pi^+\pi^-$ & Signal    & \G\   & \G\   & \GG\ \cr

955:               & \qq\ \footnote{The \qq\ \pdfs\ are the same for all modes. For brevity we omit this line in subsequent entries.}     & \ARG\ & \LIN\  & \FI \cr

956:              & Cross-feed & \G\   & \G\   & \GG\ \cr

957: $\pi^+\pi^0$ & Signal    & \GG\   & \DGG\   & \GG\ \cr

958:              & Cross-feed & \G\   & \G\   & \GG\ \cr

959: $\pi^0\pi^0$ & Signal    & \GG\   & \DGG\   & \GG\ \cr

960:              & Cross-feed & \GG\   & \GG\   & \GG\ \cr

961: \hline

962: $K^+\pi^-$   & Signal    & \G\   & \G\   & \GG\ \cr

963: $\kz\pi^+$   & Signal    & \GG\   & \DG   & \GG\ \cr

964: $K^+\pi^0$   & Signal    & \GG\   & \DGG   & \GG\ \cr

965: $\kz\pi^0$   & Signal    & \G\   & \G\   & \GG\ \cr

966: \hline

967: $K^+K^-$     & Signal    & \G\   & \G\   & \GG\ \cr

968:              & Cross-feed & \G\   & \G\   & \GG\ \cr

969: $\kz K^-$     & Signal    & \GG\   & \DG\   & \GG\ \cr

970: $\kz\kzb$     & Signal    & \GG\   & \DG\   & \GG\ \cr

971: \hline

972: $p\bar p$    & Signal    & \GG\   & \G\   & \DG\ \cr

973: $\pL$        & Signal    & \GG\   & \G\   & \DG\ \cr

974: $\LL$        & Signal    & \GG\   & \G\   & \DG\footnote{Set A includes \BW.} \cr

975: \hline

976: $D^0\pi^-$   & Signal    & \GG\   & \G\   & \GG\ \cr

977: $D^0 K^-$    & Signal    & \GG\   & \G\   & \GG\ \cr

978:              & Cross-feed & \GG\   & \G\   & \GG\ \cr

979: \hline

980: \end{tabular}

981: \label{tab:allfits}

982: \end{center}

983: \end{table}

984: 

985: 

986: %---------------------------------------------------------------------

987: %

988: \subsection{Fit Results}

989: %

990: %---------------------------------------------------------------------

991: Table \ref{tab:results} shows the results of the fits to the CLEO III

992: data.  All errors shown are statistical only.  The apparently

993: large yields of

994: \qq\ background reflect the large 

995: background-normalizing sidebands in \mb\ and \de\ and are

996: not indicative of $S/B$. Typically $S/B\sim 1$ for

997: the observed modes.

998: %In a few low-yield cases

999: %the yield was constrained to be non-negative so the fit would converge

1000: %properly; these modes are noted in the table. CHECK (For low yield cases

1001: %in which the fit converged with allowing negative yields permitted, we

1002: %checked the fit with the non-negativity constraint and found agreement

1003: %between the two methods.)

1004: 

1005: 

1006: 

1007: \begin{table}

1008: \begin{center}

1009: \caption{Results of likelihood fits: raw event yields with statistical errors.

1010: A dash in the last column means no cross-feed term was used in the fit.}

1011: \smallskip

1012: \begin{tabular}{lccccc}

1013: \hline

1014:  Mode& Set  & Eff (\%)  & Signal & \qq\ Bkg & cross-feed \cr 

1015: \hline\hline 

1016: $\pi^+\pi^-$ & A:  &  39.0 &	$7.8^{+5.5}_{-4.5}$ &	1750$\pm$42 & $3.9^{+4.6}_{-3.7}$ \\

1017:              & B:  &  45.3 &	$4.3^{+4.1}_{-3.1}$ &	1955$\pm$44 & $2.8^{+3.4}_{-2.3}$ \\

1018: 	         

1019: $\pi^+\pi^0$ & A:  &  34.9 &	$2.8^{+3.3}_{-1.9}$ &	1158$\pm$34 &	$9.3\pm 7.0$ \\

1020: 	         & B:  &  37.5 &	5.7$\pm$5.9 &		1139$\pm$34 &	$0.0\pm 3.5$ \\

1021: 

1022: $\pi^0\pi^0$ & A:  &  22.1 &	$2.2^{+2.5}_{-1.5}$ &	134$\pm$12 & $3.6^{+2.3}_{-3.1}$ \\

1023: 	         & B:  &  22.4 &	$0.4^{+2.7}_{-3.4}$ &	211$\pm$15 & $0.5^{+1.7}_{-2.6}$ \\

1024: \hline

1025: $K^+\pi^-$ & A:  &  37.9 & 	$28.1^{+6.8}_{-6.0}$ &	1779$\pm$42 &	$-$ \\

1026: 	       & B:  &  45.3 &	$19.1^{+5.3}_{-4.6}$ &	1848$\pm$43 &	$-$ \\

1027: 

1028: $\kz\pi^+$ & A:  &  12.3 &	$12.1^{+4.4}_{-3.7}$ &	398$\pm$20 &	$-$ \\

1029: 	       & B:  &  12.8 &	$2.9^{+1.8}_{-2.7}$ &	395$\pm$20 &	$-$ \\

1030: 

1031: $K^+\pi^0$ & A:  &  32.6 &	$16.7^{+6.2}_{-5.3}$ &	735$\pm$27 &	$-$ \\

1032: 	       & B:  &  35.3 &	$10.8^{+5.1}_{-4.1}$ &	780$\pm$28 &	$-$ \\

1033: 

1034: $\kz\pi^0$ & A:  &  9.6 &	$3.5^{+2.8}_{-1.9}$ &	154$\pm$13 &	$-$ \\

1035: 	       & B:  &  10.5 &	$2.9^{+2.4}_{-1.6}$ &	132$\pm$12 &	$-$ \\

1036: \hline

1037: $K^+K^-$ & A:  &  35.2 &	$2.3^{+3.7}_{-2.9}$ &	945$\pm$31 & $7.2^{+4.2}_{-3.4}$ \\

1038: 	     & B:  &  42.1 &	0.0$\pm$0.7 &		931$\pm$30 & $2.0^{+2.6}_{-1.6}$ \\

1039: 

1040: $\kz K^-$ & A:  &  13.8 &	0.0$\pm$1.5 &		371$\pm$19 &	$-$ \\

1041: 	     & B:  &  13.0 &	0.0$\pm$0.6 &		369$\pm$19 &	$-$ \\

1042: 

1043: $\kz\kzb$ & A:  &  8.1  &	0.0$\pm$0.5 &		34$\pm$6 &	$-$ \\

1044: 	     & B:  &  8.0  &	0.0$\pm$0.5 &		37$\pm$6 &	$-$ \\

1045: \hline

1046: $p\bar p$& A:  &  31.5 &	0.0$\pm$0.7 &		38$\pm$6 & $-$ \\

1047: 	     & B:  &  34.3 &$0.0^{+0.5}_{-0.6}$ &	18$\pm$5& $-$ \\

1048: 

1049: $\pL$    & A:  &  21.9 &$0.2^{+1.5}_{-0.8}$ &   46$\pm$7 & $-$ \\

1050: 	     & B:  &  21.6 &	0.0$\pm$0.8 &		44$\pm$7 &	$-$ \\

1051: 

1052: $\LL$    & A:  &  14.3 &	0.0$\pm$0.7 &		25$\pm$5 &	$-$ \\

1053: 	     & B:  &  13.4 &	0.0$\pm$0.6 &		29$\pm$6 &	$-$ \\

1054: 

1055: 

1056: \hline

1057: \end{tabular}

1058: \label{tab:results}

1059: \end{center}

1060: \end{table}

1061: 

1062: 

1063: %---------------------------------------------------------------------

1064: %

1065: \section{Systematic Uncertainties}\label{sec:syst}

1066: %

1067: %---------------------------------------------------------------------

1068: The net uncertainty in our branching ratio determinations

1069: is dominated by the statistical errors in the event yields but also

1070: includes a systematic contribution.

1071: 

1072: We categorize systematic uncertainties in two groups, multiplicative

1073: and additive.  Additive uncertainties are those that affect the overall

1074: yield of signal events, while multiplicative are those that enter as

1075: scale factors in converting the yield to a branching ratio. In view of

1076: the following equation,

1077: \begin{equation}

1078: \branch(B\to X) = \frac{N_X^{\rm observed}}{\nbb \times 

1079: {\rm (eff)} \times{\rm (secondary~BR)}}

1080: \label{eq:BR}

1081: \end{equation}

1082: the multiplicative uncertainties correspond to the uncertainty in our

1083: knowledge of the absolute number of \bbbar\ pairs in the data sample,

1084: denoted \nbb, and the reconstruction efficiency of each mode. In

1085: practice the uncertainties in the secondary branching ratios of

1086: $\pi^0\to\gamma\gamma$, $\kz\to \ks\to\pi^+\pi^-$, $\Lambda\to p\pi$ and

1087: $D^0\to (K\pi, K\pi\pz, K\pi\pi\pi)$ are negligibly small compared to

1088: uncertainties in \nbb\ and reconstruction efficiency.

1089: 

1090: 

1091: %---------------------------------------------------------------------

1092: %

1093: \subsection{Additive Systematic Uncertainties}

1094: %

1095: %---------------------------------------------------------------------

1096: 

1097: The accuracy of the signal yield obtained from the likelihood fit

1098: depends primarily on the fidelity of the \pdfs\ used in the fit. A

1099: secondary consideration is the correctness of the product form assumed

1100: in Eq. \ref{eq:evtprob}, which ignores any correlations among the four

1101: fit variable distributions. Such correlations however are expected to be

1102: small, and  Monte Carlo tests of the fit procedure confirm this

1103: expectation. We therefore focus on the systematic uncertainties in

1104: signal yield which arise from systematic uncertainties in the \pdf\

1105: parametrizations already noted in Section \ref{sec:pdfs}.  To evaluate

1106: these uncertainties we refit the data multiple times with one \pdf\

1107: parameter varied each time. The resulting signal variations are summed

1108: in quadrature, separately for negative and positive yield variations, 

1109: ignoring any correlations which may exist among the

1110: parameters.  A representative set of these uncertainties are

1111: displayed in Table \ref{tab:syst} for the \Kpi\ mode; details will vary from

1112: mode to mode. (For simplicity of presentation we have combined results

1113: from Set A and Set B, and merged the three component terms of the Fisher

1114: \pdf.) The essential feature, however, is that the net additive systematic

1115: error corresponds to a relative error of 3.5\% which is substantially

1116: smaller than that statistical error, and also smaller than the

1117: multiplicative systematic errors to be discussed next. This pattern

1118: holds true for all modes.

1119: 

1120: 

1121: \begin{table}

1122: \begin{center}

1123: \caption{Additive systematic errors due to \pdf\ variations for $B\to K^+\pi^-$.

1124: Entries show change in efficiency-corrected signal yield (events) 

1125: resulting from a parameter variation of one standard deviation up (high) or down (low).

1126: $L$ and $R$ refer to left and right sides of an asymmetric Gaussian distribution.}

1127: \smallskip

1128: \begin{tabular}{|c|l|c|c|c|}

1129: \hline

1130: \multicolumn{3}{|c|}{Parameter}&\multicolumn{2}{c|}{Result of Parameter Variation}\cr

1131: \multicolumn{3}{|c|}{~}& Low-variation & High-variation \cr

1132: \hline

1133: \multirow{4}{5mm}{\rotatebox{90}{~~~~~~~~~Signal~~~~~~~~}}&{\mb}&mean     &$-0.1$&$-0.1$\cr

1134:                                                 &               & width &$-1.4$ &$+1.3$ \cr

1135:                                                 &{\de}          &mean   &$-1.3$ &$+1.2$ \cr

1136:                                                 &               & width &$-2.8$ &$+2.5$ \cr

1137:                                                 &{\fisher}      &mean   &$-1.0$ &$+1.0$ \cr

1138:                                                 &               & width (L) &$-0.3$ &$+0.3$ \cr

1139:                                                 &               & width (R) &$-0.8$ &$+0.8$ \cr

1140: \hline

1141: \multirow{4}{5mm}{\rotatebox{90}{~~~~~Background~~~}}&{\mb}&$\lambda$   &$-1.1$&$+1.1$\cr

1142:                                                 &{\de}&slope     &$-0.1$&$+0.1$\cr

1143:                                                 &{\fisher}&mean  &$-0.6$&$+0.6$\cr

1144:                                                 &                & width (L) &$-0.6$ &$+0.6$ \cr

1145:                                                 &                & width (R) &$-1.0$ &$+1.0$ \cr

1146:                                                 &                & areas &$-0.8$ &$+0.8$ \cr

1147: \hline

1148:     & \multicolumn{2}{c|}{Total}                &    $-4.1$      &  $+3.8$\cr

1149: \hline

1150: \end{tabular}

1151: \label{tab:syst}

1152: \end{center}

1153: \end{table} 

1154: 

1155: 

1156: 

1157: 

1158: %---------------------------------------------------------------------

1159: %

1160: \subsection{Multiplicative Systematic Uncertainties}

1161: %

1162: %---------------------------------------------------------------------

1163: We summarize the multiplicative systematics in Table \ref{tab:mulsyst}.

1164: 

1165: The absolute number of \bbbar\ pairs in the data sample sets the scale

1166: for all branching ratios.  We determine this number by three different

1167: methods: counting decays of the type $B\to D^0\pi^-$, fitting

1168: distributions of the Fox-Wolfram\cite{fox} event shape variable \rfw,

1169: and direct computation from the run-by-run integrated luminosities, beam

1170: energies, and the shape of the \upsi\ resonance (normalized to

1171: 1.07 nb at the peak).  The \rfw\ method was used in previous CLEO II

1172: publications\cite{previous}, and in principle has excellent statistical

1173: power and small systematic uncertainties, but requires substantial off-resonance

1174: data that was not available in the first 30\% of the CLEO III running

1175: period.  Where off-resonance data is available, the $D\pi$ method and

1176: the \rfw\ method agree very well, and since the $D\pi$ method is

1177: available for all data sets we use it. The direct computation technique

1178: is used only as a check of the other methods, and is found

1179: to be in good agreement with them. In the $D\pi$ method, three

1180: secondary modes are used, $D^0\to K^-\pi^+$, $D^0\to K^-\pi^+\pi^0$, and

1181: $D^0\to K^-\pi^+\pi^-\pi^+$, and a small cross-feed from $B\to DK$ is

1182: subtracted.

1183:   

1184: To avoid the additional uncertainties

1185: implied by secondary $D^0$ branching ratios, we employ CLEO II \nbb\

1186: determinations to set the absolute scale for CLEO III:

1187: \begin{equation}

1188: \frac{{\RR(\bbbar)}_{III}}{{\RR(\bbbar)}_{II}}=

1189: \frac{\RR(D\pi)_{III}}{\RR(D\pi)_{II}} 

1190: \frac{{\epsilon}_{II}}{{\epsilon}_{III}}.

1191: \label{eq:nbb}

1192: \end{equation}

1193: Event rates per unit luminosity ($\RR$) and efficiencies ($\epsilon$) are

1194: determined separately for CLEO II (subscript $II$) and CLEO III

1195: (subscript $III$), and for each of the three secondary decay modes. 

1196: In the end the dominant limiting uncertainty in this technique is

1197: the statistical error in $D\pi$ yields.

1198: 

1199: Rare \bmeson\ decay modes involving \pz, \ks, or $\Lambda$ in the final

1200: state have additional uncertainties associated with the efficiency to

1201: reconstruct these particles.  We determine the reconstruction efficiencies

1202: in Monte Carlo (\effmc) simulation and then perform a separate determination in

1203: data (\effdata).  The total error in the ratio $\effdata/\effmc$, which

1204: includes both statistical errors and some systematic errors (such as

1205: branching ratios) is then interpreted as the systematic uncertainty in the

1206: reconstruction efficiency. For \pz\ the data determination consists of

1207: measuring the ratio

1208: \begin{equation}

1209: \effdata(\pz) \equiv 

1210: \frac{N(D^0\to K^-\pi^+\pi^0)/\branch(K^-\pi^+\pi^0)}{N(D^0\to K^-\pi^+)/\branch(K^-\pi^+)}

1211: \label{eq:pzeff}

1212: \end{equation}

1213: where we take the ratio of $D^0$ branching ratios obtained from

1214: reference \cite{pdg} to be $3.44\pm 0.22$.  We find $\effdata/\effmc =

1215: 1.00 \pm 0.08 \pm 0.02$ where the second error reflects conservative

1216: uncertainty in the Dalitz amplitudes of $D^0\to K^-\pi^+\pi^0$. A

1217: similar study was done using $\eta\to\gamma\gamma$, $\eta \rightarrow \pi^0\pi^0\pi^0$, and $\eta

1218: \rightarrow \pi^+\pi^-\pi^0$ decays. We anticipate that further study

1219: will refine the \pz\ systematic error estimates. A more precise

1220: determination of the systematic error, however,  is not called for by this

1221: analysis as any uncertainty under $\sim 20\%$ changes our

1222: signal sensitivities only marginally. \ks\ reconstruction uncertainty is

1223: determined similarly from comparing $D^+\to \ks\pi^+$ and $D^+\to

1224: K^-\pi^+\pi^+$, which yields $\effdata/\effmc = 1.01 \pm 0.07$. In the

1225: case of $\Lambda$ the comparison is of $\Lambda_c\to \Lambda \pi$ and

1226: $\Lambda_c\to p K\pi$, and we obtain $\effdata/\effmc = 0.93 \pm 0.17$.

1227: In the $\Lambda$ case, the net uncertainty is dominated by the

1228: relatively poorly known branching ratios.  In all cases the systematic

1229: uncertainties estimated by this technique are conservative (large) but

1230: still do not dominate the final total error.

1231: 

1232: \begin{table}

1233: \begin{center}

1234: \caption{Multiplicative systematic errors. Entries show the fractional

1235: change in branching ratios for each contributing source. Entries above

1236: the line affect all modes while those below only affect modes involving

1237: the corresponding particles. All values quoted are for CLEO III.}

1238: \smallskip

1239: \begin{tabular}{lc}

1240: \hline

1241: Source of uncertainty                         &  $\Delta\branch/\branch$ (\%) \cr

1242: \hline\hline

1243: Absolute number of \bbbar\ pairs              & 8\% \cr

1244: %Efficiency of \cossph\ cut                    & 3\% \cr

1245: Monte Carlo Statistics                        & 1\% \cr

1246: \hline

1247: Single Track Reconstruction Efficiency        & 1\% \cr

1248: Particle ID Efficiency per Identified Track   & 3\% \cr

1249: Single $\pi^0$ Reconstruction Efficiency      & 10\% \cr

1250: Single $\ks$ Reconstruction Efficiency        & 7\% \cr

1251: Single $\Lambda$ Reconstruction Efficiency    & 17\%\cr

1252: \hline

1253: \end{tabular}

1254: \label{tab:mulsyst}

1255: \end{center}

1256: \end{table} 

1257: 

1258: 

1259: %---------------------------------------------------------------------

1260: %

1261: \section{CLEO III Results}

1262: %

1263: %---------------------------------------------------------------------

1264: Event yields for the CLEO III data subsets A and B are given above in

1265: table \ref{tab:results}.  Because the signal efficiencies of Set A and

1266: Set B differ slightly the event yields in the two datasets do not have

1267: exactly the same meaning and are not directly comparable or summable. To

1268: obtain overall CLEO III results we express the measurements of Set A and

1269: Set B in the common language of branching ratios, $\branch =

1270: n_{sig}/({\nbb}\epsilon)$ forming the joint

1271: likelihood $\like_{\rm stat}(\branch) = \like_A(\branch) \like_B

1272: (\branch)$.  The subscript ``stat" emphasizes that this version of the

1273: likelihood function reflects only statistical features of the data.  We

1274: fold in systematic errors, which are common to Set A and Set B, by

1275: convolving the normalized statistical likelihood function

1276: $$

1277: {\widehat{\like}}_{\rm stat}(\branch) = 

1278: \frac{\like_{\rm stat}(\branch)}{\int_0^\infty \like_{\rm stat}(\beta)d\beta} 

1279: $$ 

1280: with both additive event

1281: yield uncertainties $\eta$, distributed according to an asymmetric

1282: Gaussian, $\GG(\eta)$, and multiplicative scale factor uncertainties

1283: $\rho$, distributed according to a symmetric Gaussian $G(\rho)$. The

1284: widths of these distributions have been discussed above. Formally this

1285: convolution may be written

1286: \begin{equation}

1287: \widehat{\like}(\branch)=

1288: \int_{-\infty}^{\infty}d\rho

1289: \int_{-\infty}^{\infty}d\eta

1290: ~{\widehat{\like}}_{\rm stat}

1291: \left(\frac{\nbb\branch\epsilon+\eta}{\nbb\epsilon(1+\rho)}\right) 

1292: \GG(\eta)G(\rho).

1293: \label{eq:convolution}

1294: \end{equation}

1295: \noindent For convenience the double convolution is performed

1296: by a Monte Carlo method. 

1297: 

1298: 

1299: The resulting distribution of ${\widehat{\like}}(\branch)$ is the final CLEO III likelihood

1300: function including all of the uncertainties in the measurement.  From it

1301: we find the minimum of the $-2 \ln {\cal L}$ distribution to measure our

1302: mean, and find the 1$\sigma$ intersections to determine the errors. 

1303: Since this is the total error, we unfold the systematic error by

1304: subtracting the statistical error in quadrature from the total error. We

1305: set 90\% confidence level upper limits by determining the value of

1306: \branch\ for which 

1307: $$

1308: \int_0^{\branch}\widehat{\like}(\beta)d\beta=0.90,

1309: $$

1310: and calculate

1311: significances by looking at the zero yield value of the $-2 \ln {\cal

1312: L}$ distribution. In the limit of a purely Gaussian likelihood function,

1313: this definition of significance reduces to the signal yield

1314: divided by its one standard deviation error.

1315: %The resulting

1316: %distributions of $-2\ln\like(\branch)$ are plotted in Figs.

1317: %\ref{fig:pipi-cleoiii}-\ref{fig:kpi-cleoiii} for the three $\pi\pi$ and

1318: %four $K\pi$ modes, and branching ratios (or upper limits) for all modes

1319: %are summarized in Table \ref{tab:results}. These results are for CLEO

1320: %III data only; in the next section we combine CLEO II measurements with

1321: %CLEO III measurements to obtain the final results.

1322: 

1323: 

1324: 

1325: %---------------------------------------------------------------------

1326: %

1327: \section{Combined CLEO II and CLEO III Results}

1328: %

1329: %---------------------------------------------------------------------

1330: We combine CLEO II and CLEO III measurements using the likelihood

1331: functions described above for CLEO III and reported in Ref.

1332: \cite{lastkpipaper} for CLEO II. For some modes we use previously

1333: unpublished likelihood functions. The baryonic modes and the \ks\ks\

1334: mode were analyzed here with the full CLEO II data set for the first

1335: time. 

1336: 

1337: Particle identification in CLEO II was limited, and modes with potential

1338: for $K/\pi$ misidentification, such as $B\to \pi^+\pi^-$ and $B\to

1339: K^+\pi^-$, were analyzed in terms of two-dimensional likelihood

1340: functions, $\like(N_{\pi^+\pi^-}, N_{K^+\pi^-})$.  The improved $K/\pi$

1341: separation in CLEO III however permits us to treat these modes

1342: independently in the new data. To combine CLEO II and CLEO III

1343: likelihood functions, therefore, we first project the two-dimensional

1344: CLEO II functions on to one-dimensional versions, using $\like(x) = \int

1345: \like(x,y) dy$, and then express in terms of branching ratios,

1346: $\like(\branch)=\like(N_{sig}/({\nbb}\epsilon))$. Systematic errors are

1347: included following the same method as described for CLEO III above to

1348: obtain a total CLEO II likelihood function for each mode.  A final

1349: combined CLEO II and CLEO III likelihood function is then formed from

1350: the joint likelihood, $\like_{\rm final}(\branch) = \like_{\rm CLEO

1351: II}(\branch)\like_{\rm CLEO III}(\branch)$. For $\pi\pi$ and $K\pi$

1352: modes with non-zero yields we plot the negative log-likelihood functions

1353: in Fig. \ref{fig:pipi-all}. Likelihood functions for the

1354: di-baryonic modes are shown in Fig. \ref{fig:baryons}.  Table

1355: \ref{table:cleoAll} summarizes the final results, with separate entries

1356: for CLEO II results (extracted from the references and reproduced here

1357: for the convenience of the reader), CLEO III results, and the combined

1358: CLEO II and CLEO III results.

1359: 

1360: \begin{figure*}[htbp]

1361: \begin{center}

1362: \mbox{

1363: \includegraphics*[width=6.0in]{Fig1.ps}  %{/nfs/web/restricted/figures/puboff/2003/2860103-001.ps}

1364: }

1365: \end{center}

1366: \caption{$-2 \ln ({\cal L}/{\cal L}_{max})$ distributions for CLEO II and CLEO III combined,

1367: for $K\pi$ and $\pi\pi$ modes with non-zero yield.} 

1368: \label{fig:pipi-all}

1369: \end{figure*}

1370: 

1371: \begin{figure}[htbp]

1372: \begin{center}

1373: \includegraphics*[width=3.0in]{Fig2.ps}  %{/nfs/web/restricted/figures/puboff/2003/2860103-002.ps}

1374: \end{center}

1375: \caption{Likelihood functions for $B\to\ppbar$, $B\to \pL$, and $B\to\LL$.} 

1376: \label{fig:baryons}

1377: \end{figure}

1378: 

1379: 

1380: \begin{table*}[htbp]

1381: \begin{center}

1382: \caption{Experimental results for CLEO II, CLEO III, and

1383: both datasets combined.  Significances include systematic errors.

1384: Note that the $p{\bar p}$ analysis in Ref. 4 was 

1385: done in only a subset of the full CLEO II dataset, so the 

1386: ``combined'' result is simply the CLEO III upper limit.  Upper limits

1387: are 90\% confidence level. CLEO II results are taken from Ref. 4,

1388: except for the $\kz\kzb$, \pL\, and \LL\ modes which were analyzed

1389: in this work with the full CLEO II dataset for the first time.}

1390: \smallskip

1391: \begin{tabular}{|l|cc|cc|cc|}

1392: \hline

1393:  & \multicolumn{2}{|c|}{CLEO II - Ref. 4}

1394:  & \multicolumn{2}{|c|}{CLEO III} 

1395:  & \multicolumn{2}{|c|}{Combined}\\

1396: \hline

1397: Mode & Significance & $\branch\power{6}$

1398:      & Significance & $\branch\power{6}$

1399:      & Significance & $\branch\power{6}$ \\ \hline\hline

1400: $\pi^+\pi^-$ & 4.2    & 4.3$^{+1.6+0.5}_{-1.4-0.5}$ 

1401:              & 2.6    & 4.8$^{+2.5+0.8}_{-2.2-0.5}$ 

1402:              & 4.4    & 4.5$^{+1.4+0.5}_{-1.2-0.4}$ \\

1403: $\pi^+\pi^0$ & 3.2    & 5.6$^{+2.6+1.7}_{-2.3-1.7}$  

1404:              & 2.1    & 3.4$^{+2.8+0.8}_{-2.0-0.3}$ 

1405:              & 3.5    & 4.6$^{+1.8+0.6}_{-1.6-0.7}$ \\

1406: $\pi^0\pi^0$ & 2.0    & $(<5.7)$                    

1407:              & 1.8    & $(<7.6)$                    

1408:              & 2.5    & $(<4.4)$                    \\ \hline

1409: $K^+\pi^-$   & $12$   & 17.2$^{+2.5+1.2}_{-2.4-1.2}$

1410:              & $ >7$  & 19.5$^{+3.5+2.5}_{-3.7-1.6}$

1411:              & $>7$   & 18.0$^{+2.3+1.2}_{-2.1-0.9}$\\

1412: $\kz\pi^+$   & 7.6    & 18.2$^{+4.6+1.6}_{-4.0-1.6}$

1413:              & 4.6    & 20.5$^{+7.1+3.0}_{-5.9-2.1}$

1414:              & $>7$   & 18.8$^{+3.7+2.1}_{-3.3-1.8}$\\

1415: $K^+\pi^0$   & $6.1$  & 11.6$^{+3.0+1.4}_{-2.7-1.3}$

1416:              & $5.0$  & 13.5$^{+4.0+2.4}_{-3.5-1.5}$

1417:              & $>7$   & 12.9$^{+2.4+1.2}_{-2.2-1.1}$\\

1418: $\kz\pi^0$   & 4.9    & 14.6$^{+5.9+2.4}_{-5.1-3.3}$  

1419:              & 3.8    & 11.0$^{+6.1}_{-4.6}\pm2.5$  

1420:              & 5.0    & 12.8$^{+4.0+1.7}_{-3.3-1.4}$\\ \hline

1421: $K^+K^-$     & -      & $(<1.9)$                    

1422:              & -      & $(<3.0)$                    

1423:              & -      & $(<0.8)$                    \\

1424: $\kz K^-$    & -      & $(<5.1)$                    

1425:              & -      & $(<5.0)$                    

1426:              & -      & $(<3.3)$                    \\

1427: $\kz\kzb$    & -      & $(<6.1)$                    

1428:              & -      & $(<5.2)$                    

1429:              & -      & $(<3.3)$                    \\ \hline

1430: $\ppbar$     & -      & $(<7.0)$                    

1431:              & -      & $(<1.4)$                    

1432:              & -      & $(<1.4)$                    \\

1433: $\pL$        & -      & $(<2.0)$                    

1434:              & -      & $(<3.2)$                    

1435:              & -      & $(<1.5)$                    \\

1436: $\LL$        & -      & $(<1.8)$                    

1437:              & -      & $(<4.2)$                    

1438:              & -      & $(<1.2)$                    \\ \hline

1439: \end{tabular}

1440: \label{table:cleoAll}

1441: \end{center}

1442: \end{table*}

1443: 

1444: 

1445: 

1446: 

1447: 

1448: %---------------------------------------------------------------------

1449: %

1450: \section{Physically Interesting Ratios and the Phase of $V_{ub}$}

1451: %

1452: %---------------------------------------------------------------------

1453: 

1454: As discussed in the introduction, it is possible to extract information

1455: about the phase of \vub\ from these charmless \bmeson\ decay data. The

1456: method of Reference 3 is based on two ratios of the branching fractions

1457: which we have measured and reported above.  Using the notation of this

1458: reference, and combining statistical and systematic errors

1459: in quadrature, the ratios are found to be:

1460: 

1461: \begin{equation}

1462: R_{*}\equiv

1463: \frac{\branch(B^\pm \to K^0 \pi^\pm)}{2\branch(B^\pm \to K^\pm\pi^0)}

1464: = 0.73 \pm 0.21,

1465: \end{equation}

1466: 

1467: and

1468: 

1469: \begin{equation}

1470: \epsilon_{exp}\equiv \tan\theta_C\frac{f_K}{f_\pi}

1471: \left[

1472: \frac{2\branch(B^\pm \to \pi^\pm\pi^0)}{\branch(B^\pm \to K^0 \pi^\pm)}

1473: \right]^{1/2} = 0.18 \pm 0.04.

1474: \end{equation}

1475: 

1476: We see that the precision available with the CLEO data is about 20-30\%

1477: in these quantities. With data from the {\it BABAR} and Belle

1478: experiments\cite{babarandbelle} we can make world (weighted) averages of

1479: branching ratios and reach 10-15\% experimental precision in the

1480: critical ratios: $R_{*}=0.71\pm 0.09$ and $\epsilon_{exp}=0.21\pm 0.02$.

1481:  These numbers in turn indicate a preferred region for $\gamma =

1482: Arg(\vub)$ which is greater than 90${}^\circ$\cite{alan}. Using these

1483: world-averaged data we construct contours in the $\rho-\eta$ plane

1484: according to prescription of Reference 3 and display the result in

1485: Figure \ref{fig:alan}.  The dark band represents the experimental

1486: central value convolved with theoretical uncertainties; lighter bands

1487: show the additional coverage when 68\% and 95\% experimental confidence

1488: regions are included. For reference we also overlay  68\% and 95\%

1489: confidence level ellipses of the preferred apex of the unitarity

1490: triangle as obtained in a standard analysis based on $B$ mixing, $\sin

1491: 2\beta$, \vub, and kaon decays\cite{stochhi}. An intriguing discrepancy

1492: between these regions is noticeable.  In the short term the most

1493: substantial progress to be made will be in reducing the statistical

1494: errors on the branching ratios of charmless \bmeson\ decay modes. If

1495: discrepancies survive there could be non-trivial implications for the

1496: Standard Model, as discussed in Ref. 3.

1497: 

1498: \begin{figure}[hbtp]

1499: \begin{center}

1500: \mbox{

1501: \includegraphics*[width=3.0in]{Fig3.ps}  %/nfs/web/restricted/figures/puboff/2003/2860103-004.ps}

1502: }

1503: \end{center}

1504: \caption{Confidence contours in the $\eta-\rho$ plane.  The shaded

1505: bands represent regions allowed by the world-averaged charmless \bmeson\ decay

1506: measurements while the ellipsoids represent 68\% and 95\% contours

1507: from conventional global fits to heavy quark measurements. The

1508: dark shaded region corresponds to the experimental central values

1509: of the charmless data, smeared by theoretical uncertainty.  See

1510: text for details and references.}

1511: \label{fig:alan}

1512: \end{figure}

1513: 

1514: 

1515: 

1516: 

1517: %---------------------------------------------------------------------

1518: %

1519: \section{The $\branch(B\to DK)/\branch(B\to D\pi)$ ratio}

1520: %

1521: %---------------------------------------------------------------------

1522: 

1523: \begin{figure*}[hbtp]

1524: \begin{center}

1525: \mbox{

1526: \includegraphics*[width=6.0in]{Fig4.ps} %{/nfs/web/restricted/figures/puboff/2003/2860103-003.ps}

1527: }

1528: \end{center}

1529: \caption{CLEO III data: the $M_{B}$ distribution for  $B^- \rightarrow D^0\pi^-$ 

1530: (left) and $B^- \rightarrow D^0K^-$ (right) candidates.}

1531: \label{fig:kamal}

1532: \end{figure*}

1533: 

1534: 

1535: In view of the good $K/\pi$ separation in CLEO III data we also report a

1536: new determination of the ratio $\branch(B^-\to D^0K^-)/\branch(B^-\to

1537: D^0\pi^-)$ which benefits substantially from good particle

1538: identification. The original CLEO II publication is available

1539: in Ref. \cite{abi}.

1540: 

1541: For this analysis, $D^0$ candidates are reconstructed in three secondary

1542: modes, $D^0\to K^-\pi^+$, $D^0\to K^-\pi^+\pi^0$, and $D^0\to

1543: K^-\pi^+\pi^-\pi^+$. Requirements for the $B\to D^0h^-$ modes include a

1544: 30 MeV $D^0$ mass cut, a 100 MeV \de\ cut, and standard particle ID as

1545: described previously on both the primary $h^-$ from the \bmeson\ and on

1546: the secondary kaon from the $D^0$.  The $\pi^0$ mass for the $D^0 \ra

1547: K^-\pi^+\pi^0$ mode is required to be within 30 MeV of its nominal

1548: mass. For the $D^0 K$ likelihood fit, $D^0 \pi$ is included as a

1549: cross-feed background, and  corresponds to 

1550: approximately 50\% of the DK yield shown in Fig. \ref{fig:kamal}.

1551: (Both $D^* \pi$ and $D\rho$ were found not to be

1552: significant backgrounds to either signal mode.) Fig. \ref{fig:kamal}

1553: shows \mb\ distributions for $B^-\to D^0\pi^-$ and $B^-\to D^0 K^-$ candidates

1554: with the likelihood fit shape superimposed.

1555: 

1556: 

1557: Combining the three $D^0$ submodes, we find

1558: \begin{equation}

1559: \frac{\branch(B^-\ra D^0 K^-)}{\branch(B^-\ra D^0 \pi^-)} = 

1560: (9.9^{+1.4+0.7}_{-1.2-0.6})\power{-2}.

1561: \end{equation}

1562: Most systematic errors cancel in this ratio, with only a small

1563: residual arising from the particle identification requirements

1564: imposed on the primary $\pi/K$ in both numerator and denominator.

1565: 

1566: 

1567: 

1568: %---------------------------------------------------------------------

1569: %

1570: \section{Summary}

1571: %

1572: %---------------------------------------------------------------------

1573: We have presented final results from the CLEO experiment on charmless hadronic

1574: \bmeson\ decays. The decay modes include the ten \pipi, \Kpi, and \kk\

1575: final states as well as the dibaryonic states $p\bar p$, $p\bar \Lambda$

1576: and \LL.  In addition we have presented a new determination of the ratio

1577: of branching ratios $\branch(B\to DK)/\branch(B\to D\pi)$. The results

1578: are based on the full CLEO II and CLEO III data samples totalling

1579: 15.3 \fbinv\ at the $\Upsilon(4S)$, and supercede previously

1580: published results by this collaboration.

1581: 

1582: 

1583: 

1584: %---------------------------------------------------------------------

1585: %

1586: \section{Acknowledgements}

1587: %

1588: %---------------------------------------------------------------------

1589: We gratefully acknowledge the effort of the CESR staff in providing us

1590: with excellent luminosity and running conditions. M. Selen thanks the

1591: Research Corporation, and A.H. Mahmood thanks the Texas Advanced

1592: Research Program. This work was supported by the National Science

1593: Foundation, and the U.S. Department of Energy.

1594: 

1595: 

1596: 

1597: %---------------------------------------------------------------------

1598: %

1599: %\section{Figures}

1600: %

1601: %---------------------------------------------------------------------

1602: 

1603: 

1604: 

1605: %---------------------------------------------------------------------

1606: %

1607: %\section{Bibliography}

1608: %

1609: %---------------------------------------------------------------------

1610: \def\endpoint{;~~}

1611: \def\Journal#1&#2&#3(#4){#1{\bf #2}, #3 (#4)}

1612: \def\NIM{Nucl. Instr. and Meth. }

1613: \def\NIMA{Nucl. Instr. and Meth. A }

1614: \def\NPB{Nucl. Phys. B }

1615: \def\PLB{Phys. Lett. B }

1616: \def\PRL{Phys. Rev. Lett. }

1617: \def\PRD{Phys. Rev. D }

1618: \newpage

1619: \begin{thebibliography}{99}

1620: 

1621: \bibitem{fleischeretal} 

1622: Y.-Y. Keum, H.-N. Li, and A.I. Sanda, arXiv:hep-ph/0201103; 

1623: M. Neubert,                  {JHEP} {\bf 9902} (1999) 014;

1624: M. Neubert and J.L. Rosner,  {\Journal\PRL&81&5076(1998)};

1625: M. Neubert and J.L. Rosner,  {\Journal\PLB&441&403(1998)};

1626: M. Neubert,                  {\Journal\PLB&424&2752(1998)};

1627: R. Fleischer and T. Mannel,  {\Journal\PRD&57&2752(1998)}

1628: 

1629: \bibitem{DK} M. Gronau and D. Wyler, {\Journal\PLB&265&172(1991)}

1630: 

1631: \bibitem{neubertnew}\label{ref:neubert}

1632: M. Neubert, ArXiv:hep-ph/0207327, CLNS-02/1794\endpoint  

1633: M. Beneke, G. Buchalla, M. Neubert, and C.T. Sachrajda, {\Journal\NPB&606&245(2001)}\endpoint

1634: M. Beneke, G. Buchalla, M. Neubert, and C.T. Sachrajda, {\Journal\PRL&83&1914(1999)}\endpoint  

1635: 

1636: \bibitem{previous} \label{ref:previous}

1637: CLEO publications on charmless hadronic B decays:\newline

1638: (a) D. M. Asner \etal,       {\Journal\PRD&65&031103(2002)} \endpoint % pi0pi0 %D. Asner \etal,  CLEO Collaboration, Cornell Report No. CLNS 01/1718, CLEO 01-02  (2001)\endpoint 

1639: (b) {R.A. Briere \etal, }    {\Journal\PRL&86&3718(2001)}\endpoint % phi K, phi K*

1640: (c) D. Cronin-Hennessy \etal,{\Journal\PRL&85&515(2000)}\endpoint % pi+pi-, K0pi0

1641: (d) {S.J. Richichi \etal, }  {\Journal\PRL&85&520(2000)}\endpoint % eta K*

1642: (e) {S. Chen \etal, }        {\Journal\PRL&85&525(2000)}\endpoint % cp asyms

1643: (f) {C. Jessop \etal, }      {\Journal\PRL&85&2881(2000)}\endpoint % PV modes

1644: (g) {R. Godang \etal, }      {\Journal\PRL&1705&26(2000)}\endpoint % VV modes

1645: (h) {T.E. Coan \etal, }      {\Journal\PRD&59&111101(1999)}\endpoint % curt ward baryonic search

1646: (i) {B.H. Behrens \etal, }   {\Journal\PRL&80&3710(1998)}\endpoint % decays to eta, eta'

1647: (j) {R.~Godang\etal, }       {\Journal\PRL&80&3456(1998)}\endpoint % first obs Kpi

1648: (k) {T. Bergfeld \etal, }    {\Journal\PRL&81&272(1998)} \endpoint % omega K

1649: (l) {T.E. Browder \etal, }   {\Journal\PRL&81&1786(1998)}\endpoint % inclusive eta'

1650: (m) {D.M. Asner \etal, }     {\Journal\PRD&53&1039(1996)}\endpoint % big rare b 

1651: (n) {M. Battle \etal, }      {\Journal\PRL&71&3922(1993)}\endpoint % first kpi+pipi paper

1652: (p) {C. Bebek \etal, }       {\Journal\PRL&62&8(1989)}   \endpoint  % no ppbar pi

1653: (q) {D. Bortoletto \etal, }  {\Journal\PRL&62&2436(1989)}\endpoint % jim Mueller

1654: (r) {P. Avery \etal, }       {\Journal\PLB&183&429(1987)} % search for exclusive penguin

1655: 

1656: \bibitem{babarandbelle} 

1657:  B. Aubert, \etal,  BABAR Collaboration, ArXiv:hep-ex/0207055;

1658:  B. Aubert, \etal,  BABAR Collaboration, ArXiv:hep-ex/0207063;

1659:  B. Aubert, \etal,  BABAR Collaboration, ArXiv:hep-ex/0207065;

1660:  B. Aubert, \etal,  BABAR Collaboration, ArXiv:hep-ex/0206053;

1661:  B.C.K. Casey, \etal,  Belle Collaboration, {\Journal\PRD&66&092002(2002)};

1662:  K. Abe, \etal,        Belle Collaboration, {\Journal\PRD&66&092002(2002)} 

1663:  

1664: \bibitem{cleo3det}     CLEO Collaboration, CLNS-94-1277; D. Peterson \etal, {\Journal\NIMA&478&142(2002)}

1665: 

1666: \bibitem{cleo2det} Y.~Kubota {\etal, } (CLEO Collaboration), {\Journal\NIMA&320&66(1992)};  

1667:                    T.S.~Hill,                                {\Journal\NIMA&418&32(1998)}.

1668: \bibitem{rich} 

1669: M. Artuso \etal,  "Construction, Pattern Recognition and Performance of the CLEO III LiF-TEA RICH Detector," Presented at Fourth Workshop on RICH Detectors, Pylos Greece, June, 2002, to appear in the proceedings [hep-ex/0209009]

1670: 

1671: \bibitem{pdg} Particle Data Group, {\Journal\PRD&66&010001(2002)}.\label{ref:pdg}

1672: 

1673: \bibitem{fplusminuscleo} {J.P. Alexander \etal, } (CLEO Collaboration)  {\Journal\PRL&86&2737(2001)}

1674: 

1675: \bibitem{Fisher} R.A. Fisher, {\it The Use of Multiple Measurements in Taxonomic Problems}, Annals of Eugenics, 7, (1936) 179.

1676: 

1677: \bibitem{bigrareb} D.~M.~Asner \etal,  (CLEO Collaboration), {\Journal\PRD&53&1039(1996)}.

1678: 

1679: \bibitem{argus} ARGUS Collaboration, H.~Albrecht \etal,  {\Journal\PLB&241&278(1990)}; {\Journal\PLB&254&288(1991)}.

1680: 

1681: \bibitem{fox} G.~Fox and S.~Wolfram, {\Journal\PRL&241&1581(1978)}, 

1682: 

1683: \bibitem{lastkpipaper}\label{ref:lastkpipaper} 

1684:         D. Cronin-Hennessy \etal, (CLEO Collaboration), {\Journal\PRL&285&515(2000)} ; 

1685: \bibitem{alan} Alan J. Magerkurth, Cornell University Ph.D thesis (unpublished)

1686: {\it Measurement of Two Body B Meson Decays to Pions and Kaons with the CLEO III Detector}.

1687: 

1688: 

1689: \bibitem{stochhi}  A. Stocchi, {\it Proceedings of the XXXIst Int'l Conf. on

1690: HEP}, Amsterdam 2002. LAL 02-102; arXiv:hep-ph/0211245

1691: 

1692: \bibitem{abi} {M. Athanas \etal, }  (CLEO Collaboration) {\Journal\PRL&80&5493(1998)}

1693: 

1694: \end{thebibliography}

1695: 

1696: 

1697: \end{document}

1698: